1. Introduction
Leukemias and lymphomas are the most commonly diagnosed types of childhood cancer worldwide [
1], and are responsible for the most cancer deaths among children [
1]. Thus, underutilized avenues of care must be exploited to improve the survival of children with leukemia and lymphoma. Nutrition is a documented determinant of clinical outcomes for pediatric cancer patients. As many as 46% of children and young adults receiving treatment for cancer experience malnutrition [
2], and several studies suggest that malnutrition portends worse outcomes for children with cancer. For example, under nutrition at diagnosis is associated with decreased 5-year survival [
3] and inferior event free survival [
4] of children with acute leukemia. Malnourishment at diagnosis has been associated with an increase in treatment related neutropenia in children being treated for Burkitt lymphoma [
5]. Conversely, excessive body weight caused by over nutrition is associated with an increase in risk for relapse in children treated for acute lymphoblastic leukemia [
6]. Overweight and obesity can also exacerbate iatrogenic effects of treatment including heart disease and osteoporosis as pediatric survivors transition into adulthood [
7]. Thus, although there is strong evidence that nutrition has prognostic significance for pediatric patients with leukemia and lymphoma, few studies have attempted to clarify the molecular basis for this observation.
Cellular oxidative stress is associated with both dietary intake and cancer therapy. Oxidative stress represents a state of unbalance between reactive oxygen species (ROS) and cellular antioxidants. It has been demonstrated that dietary intake alters or modifies ROS levels in healthy populations. Meydani et al. [
8] reported that restricting overweight adults’ calorie consumption for 6 months increased the plasma levels of the antioxidant enzyme glutathione peroxidase and decreased the plasma levels of protein carbonyls, which suggest that reduced calorie consumption may reduce oxidative stress. Dietary phytochemicals have also been explored as potential regulators of oxidative stress in the context of cancer prevention and treatment [
9].
While oxidative stress under normal conditions may lead to the progression of disease or stress-induced DNA damage, increased levels of intracellular oxidative stress are thought to contribute to the efficacy of several chemotherapeutic regimens utilized in pediatric leukemia and lymphoma patients. These chemotherapeutic agents capitalize on the cellular redox system to increase ROS in cancer cells to the point of irreparable damage. Increased cellular oxidative stress during cancer treatment has been previously shown in pediatric patients. Papageorgiou et al. [
10] demonstrated total antioxidant capacity of plasma from children with various malignancies was reduced during chemotherapy compared to capacity at diagnosis and after treatment. However, this study used total antioxidant capacity, a combination of endogenous and dietary anti-oxidants, as opposed to direct measures of oxidative stress, and did not explore factors that may influence redox.
Associations between dietary compounds and oxidative stress during chemotherapy may reveal some molecular basis for the impact of nutrition during treatment on patient outcomes. Since many chemotherapeutics used for the treatment of pediatric malignancies increase levels of reactive oxygen species (ROS) [
11] and because diet may modulate the balance of cellular pro- and antioxidants, we sought to correlate the nutritional status of pediatric cancer patients receiving therapy with changes in direct measures of oxidative stress. Childhood leukemia and lymphoma patients were followed for six months during chemotherapy regimens. Oxidative stress increased over this time frame and was associated with protein consumption.
2. Materials and Methods
2.1. Participants and Procedure
The Institutional Review Board of the University of Texas MD Anderson Cancer Center approved this protocol (2009-0954). This was a prospective cohort pilot study. Eligible participants were seven years old or younger who were undergoing treatment for leukemia or lymphoma at the MD Anderson Children’s Cancer Hospital. Patients were excluded from the study if they were already enrolled on a dietary intervention. Participants were identified as potentially eligible for this study by their attending physician or study staff. Recruitment was conducted in the clinic by examining scheduled appointments weekly, and eligible patients were approached for study participation during scheduled appointments. Participation was completely voluntary, and was approved by the patient’s attending pediatric oncologist. Parents of all participants completed an informed consent or provided assent prior to data collection.
This study utilized peripheral blood samples to assess oxidative stress in mononuclear cells isolated from participating pediatric cancer patients. Dietary recalls were conducted at time points coinciding with blood draws. Baseline measures were captured when participants were off chemotherapy for at least two weeks but not prior to initiation of treatment, and participants were followed for six months with additional data collection at baseline, one month, two months and six months as shown in
Figure 1.
2.2. Cells
Peripheral blood mononuclear cells (PBMCs) were isolated as follows. Five milliliters of whole blood, drawn from study participants into heparin-containing vaccutainer tubes, were slowly layered on top of an equal volume of Ficoll-PaqueTM PLUS (GE Healthcare Humble, TX, USA) and centrifuged at 1500 rpm for 30 min at room temperature. The fraction containing PBMCs was isolated, washed twice in phosphate buffered saline (PBS), and counted with a hemocytometer after staining with trypan blue. The Jurkat T-cell leukemia line was purchased from the American Type Culture Collection (Manassas, VA, USA). They were grown in a humidified incubator with 5% CO2 at 37 °C and were cultured in RPMI 1640 with 10% (v/v) heat-inactivated fetal bovine serum (Hyclone, Logan, UT, USA), 2 mM L-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin (Sigma, St. Louis, MO, USA).
2.3. Quantification of ROS
Intracellular superoxide and hydrogen peroxide levels were measured using dihydroethidium [
12] (HE) (Molecular Probes/Invitrogen, Carlsbad, CA, USA) and 5-(and-6)-chloromethyl-2′,7′-dichlorodihydrofluorescein diacetate [
13], acetyl ester (DCF-DA) (Molecular Probes/Invitrogen, Carlsbad, CA, USA), respectively. Briefly, 5 × 105 PBMCs were resuspended in 1 mL PBS containing 10 µM HE or 1 mL RPMI containing 10 µM DCF and incubated for 30 min at 37 °C in the dark. Cells were then washed twice with PBS and read on the FL-1 (for hydrogen peroxide) or FL-3 (for superoxide) channel of a Becton Dickinson fluorescence-activated cell sorter (FACS Caliber; Becton-Dickinson, Franklin Lakes, NJ, USA). Jurkat cells (5 × 105) were treated with 2 mM hydrogen peroxide (Sigma, St. Louis, MO, USA) for 15 min and stained as above as a positive control for ROS staining. Results were analyzed using Cell Quest Software (Becton Dickinson). Mean fluorescence for each sample was calculated by subtracting baseline fluorescence (unstained cells) then normalizing to the cell number.
2.4. Quantification of Glutathione
Intracellular glutathione levels were measured [
14]. Briefly, 1 × 10
6 PBMCs were resuspended in 1 mL of a 50 mM monochlorobimane (Sigma, St. Louis, MO, USA) solution, and incubated for 15 min at 37 °C in the dark. After incubation, the stain was quenched by adding 50 µL of trichloroacetic acid to each sample and samples were centrifuged at 10,000 rpm for 5 min at room temperature. One milliliter of the supernatant was then layered on top of an equal volume of dichloromethane in a glass tube and the sample was centrifuged at 3500 rpm for 2 min at room temperature. The aqueous supernatant (200 uL/well) was plated in duplicate in an opaque 96-well plate and the fluorescence was read immediately on a Spectra GemniEM plate reader at excitation 360/40 nm and emission 460/40 nm. Standards of known concentrations of reduced glutathione were prepared in parallel with each sample and were used to calculate glutathione concentration. Jurkat cells (1 × 10
6) were treated with 2 mM glutathione ethylester (Sigma, St. Louis, MO, USA) for 15 min or 2 mM hydrogen peroxide (Sigma, St. Louis, MO, USA) as a positive control and negative control, respectively. Results were analyzed using Microsoft Excel 2010. Mean fluorescence for each sample was calculated by subtracting baseline fluorescence (unstained cells) then normalizing it to the cell number.
2.5. Nutrition Measurements
Twenty-four hour dietary recalls were completed at each assessment to determine the participants’ habitual food intake. Study staff met participants’ guardians in person at the MD Anderson Children’s Cancer Hospital to complete the dietary recalls. Participants’ guardians were asked to recall the foods consumed by the participant during the previous 24 h, including details about the timing of meals, food preparation, and portion sizes. Guardians were also provided with photographs and measuring supplies to aid in the accurate recall of portion sizes when needed. The consumption of specific micronutrients was measured by analyzing the dietary intake information using the University of Minnesota Nutrition Data System for Research (NDSR) Software, version 2011.
2.6. Demographic and Anthropometric Measurements, and Assessment of Blood Counts
We examined participant medical records to record basic demographic information as well as changes in height and weight during the course of our study. Height and weight were recorded on the days that blood samples were collected and processed for markers of oxidative stress. Body mass index (kg/m
2) was calculated using the height and weight. Participant medical records were also examined to determine white blood cell counts, hemoglobin levels, platelet counts, and whether patients exhibited signs of infection, such as fever. All these measures were collected at each of the visits depicted in
Figure 1.
2.7. Response to Therapy
To measure response to therapy, we examined the participant medical records at each assessment for adverse events as defined by the Common Terminology Criteria for Adverse Events, version 3. We also used the medical records to record the presence or absence of minimal residual disease (defined as greater than 0.01% leukemia blasts in bone marrow measured by flow cytometry) and absolute lymphocyte counts at day 28 or 29, two common indicators of prognosis.
2.8. Statistical Analysis
Demographic information was summarized using descriptive statistics including mean, standard deviation, median, and range for continuous variables, and frequency and proportion for categorical variables. Association between categorical variables was examined by Chi-Squared test or Fisher’s exact test when appropriate. Linear regression was used to examine differential changes over time for oxidative stress measures. Oxidative stress measures were log-transformed and fitted into a linear mixed model, which included a group effect of age, time effect of visit, and interaction effect between age and visit. Dietary data were extracted from 24 h recalls and analyzed using the Nutrition Data System for Research (NDSR). Nutrition and clinical measures were added to the model separately. All computations were carried out in SAS 9.3 (SAS Institute Inc., Cary, NC, USA).
4. Discussion
This pilot study explored the relationship between nutrition, therapy response and redox status in 32 pediatric leukemia and lymphoma patients over a six-month window of chemotherapy. We demonstrated that intracellular superoxide, peroxide and glutathione levels increased over time in the six-month period of chemotherapy treatment for the entire sample and were associated with animal protein, vegetable protein, and total protein intake. Younger children (under 48 months) and older children (over 48 months) showed significantly different baseline levels of reactive oxygen species, a measure of oxidative stress. This difference in younger versus older participants may be attributable to lower mutation rates, less exposure to environmental carcinogens, age associated epigenetic differences, or cell cycle rates, and was not addressed in our current investigations but is worthy of future study.
Changes in hydrogen peroxide and superoxide showed a positive trend in the older group, but not the younger group. This change appeared to be independent of changes in macronutrient intake as participant diet did not shift dramatically over the study period (
Table 3). These findings support previous observations that certain chemotherapeutic drugs may function through increased generation of ROS over the course of active treatment [
15,
16]. It is unclear why trends differed by age group; this finding warrants further exploration in future studies.
Glutathione was positively associated with vegetable protein. Glutathione is an abundant antioxidant with the capacity to counter oxidative stress damage within cells. Elevated glutathione levels has been positively associated with relapse of disease among pediatric ALL patients [
17]. Given that many chemotherapeutics used in ALL therapy optimize the oxidative stress pathway to cause cancer cell death, a deeper examination of the impact of diet on glutathione levels during treatment is warranted.
This finding is in line with previous research that showed differentiation between animal and vegetable protein with regard to ROS. Gu et al. found that high casein protein (derived from dairy) diets increased ROS in duodenum, liver, and pancreas of mice. However, this increase was not found in mice fed high soy protein diets [
18]. Although different ROS measures were used in this study than those described in the current work, these findings suggest vegetable as opposed to dairy protein may have different impacts on ROS generation, potentially due to glutathione’s ability to remove ROS from leukemia cells [
17].
The interplay of dietary protein sources, pro-oxidant generation, anti-oxidant generation, and cancer prognosis is still unclear. More recent work using the Eμ-Myc-lymphoma mouse demonstrated that an overall low protein diet significantly slowed cancer progression [
19]. These findings highlight the need for more nuanced research examining the mechanisms linking diet, oxidative stress, and cancer. Prospective applications of this research may include investigating the modulation of protein consumption concomitant with chemotherapy in this population. Detailed investigation into the role of specific amino acids or specific types of protein sources in augmenting chemotherapy hold potential to improve outcomes for leukemia and lymphoma patients through diet modification.
This work is unique as it examines direct measures of oxidative stress, as opposed to more commonly used measures of overall antioxidant capacity, thus, deepening our understanding of the role specific antioxidants play in redox homeostasis during therapy. Our findings demonstrate proof of principle that chemotherapy drugs increase direct measures of oxidative stress, and identify a novel association of protein intake with oxidative stress measures in pediatric tumor patients.
This study has several limitations including a relatively small sample size, which limits the generalizability of our results. The sample population was inclusive of a range of different pediatric hematological malignancies (
Table 1) and enrollment occurred at different stages of treatment. Thus, we cannot determine if all chemotherapeutic drugs impacted oxidative stress measures or only specific ones. Some of most commonly used drugs in our sample included vincristine, methotrexate, and dexamethasone. Dexamethasone is known to influence appetite and was not consistently delivered or dosed across all patients studied because of the range of diagnoses and stages. Glutathione and oxidative stress levels may impacted by transfusions and infection. Infection status was not associated with oxidative stress in our study (
Table 6); however, transfusion status was not examined. Finally, dietary recalls may be subject to recall bias and may not accurately capture a patient’s true diet. No data on nutrient absorption or data on alternative modulators of vitamin levels (i.e., vitamin D from sunlight) were collected, and this is an important variable in considering biological effects. Future research should consider examining oxidative stress relative to dietary intake variables among larger and more diverse patient populations.